Strategies of Artificial Intelligence

A hundred years ago, we started making mass-produced clothes as we know it today, rather than having everyone sew their own clothes. The tailors at the time were not worried, however, from their point of view it did not make sense that anyone would wear clothes that did not fit perfectly. What turned out to be the most important thing, though, was that for most people, mass produced clothing was good enough, considering the price.

The same issue will created by artificial intelligence (machine learning) for a multitude of industries in the near future. And just as a tailor had a very unfair fight against large factories that produced garments on assembly lines - so there will be great strategic benefits to be given to the companies that can embrace the development rather than trying to fight it.

Typically, letting computers take over some of the work processes that are delivered through consultants and service agreements today will not replace people's work from one day to the next. Instead, it will be that computer algorithms can offer new and different ways of doing the work, and where customers feel that it is good enough - the company that provides the computerized service will enjoy great growth. Elsewhere, customers will increasingly appreciate human contact and creativity, and algorithms will be inadequate.

But the question then is whether the service your company offers is one that can be performed well enough by a computer? As with tailors, there is probably always a tendency to overestimate customers' appreciation of personal service, so just like editing text is best done by someone other than the author - so it might be best to ask one who is impartial.

An industry that is currently in a panic is the industry for investment consultants. It is an industry with very high salaries, from the combination of the fact that customers have a lot of money and that at the same time they rarely know much about investments, they have made credible and skilled consultants a sought-after product. There is just the problem that various fraud cases have made customers actually rely more on a computer that works without thinking about personal gain.

That's why many rich people from Silicon Valley and elsewhere are flocking to new companies like Wealthfront and Betterment. These services place your money entirely automated in a wide variety of secure mutual funds. They do not try to pick the best stocks or otherwise predict the future (with increased risk as a consequence). They have simply automated the most common techniques for selecting a "boring" mixed portfolio that follows the market in general and minimizes taxes and fees. If the alternative is to risk being the victim of the next Bernie Madoff, then for most people this is good enough.

Another industry that will potentially be close to 100% automated is real estate agents. The US company OpenDoor uses machine learning to assist in appraising properties, so if you go and consider selling your house, they can immediately offer you a price. All you need to do is fill in some questions about the house and the condition of it. If you accept the offer and have filled out the form truthfully then in a few weeks you may have sold your house. Once the house has been sold, OpenDoor acts very much like a used car dealer and they will try to sell it as soon as possible.

There can be many reasons to want to sell your house, and getting rid of it quickly (but still at a fair price) will probably appeal to many. And even if the offer is a bit low, with the typical real estate brokerage fees there is a lot of buffer - in many cases OpenDoor's offer will probably be good enough.

There has been a lot of talk in the media lately about the upcoming self-driving cars. And it is definitely an intriguing thought that you can sit and work on your laptop on the way to work, or maybe just take a nap on the way home - it is an exciting future for us as consumers. What, however, has not been enough focus in my opinion, is what it will do for the transport industry.

In both the US and Denmark, approx. 1 percent of the workforce are employed as truck drivers. Not only will a self-driving truck be able to drive for close to 24 hours a day, it will also be able to do so with far fewer accidents on the roads. It is hard to see how logistics companies should not be compelled to be first to replace their trucks with self-driving vehicles of sorts. The day when it happens will be a major upheaval for the community. It will affect not only the many truckdrivers but also gas stations and shops along the road network.

We are right now in the machine learning golden age. Everywhere there are opportunities and threats depending on whether you are on one or the other side of the development. As a society, it will present new challenges in coping with the great upheavals.

These developments certainly do not happen from one day to the next, but it will happen faster than most people expect. And just as you are having a hard time finding a .com internet domain name that has not been taken, in just 10 years you will have a significantly harder time taking advantage of the development in machine learning that is going on right now.

So if you can see the potential for your business, now is the time to invest. Acting now and getting us into the future via a smooth transition is probably also the best way to handle some of the challenges. We should all expect our jobs to change, but you can choose to see it as a positive challenge and show the way forward.